Human parainfluenza viruses 1, 2, and 3 (HPIV1, 2, 3) are significant causes of severe pediatric respiratory tract disease worldwide. The HPIVs are enveloped, non-segmented, negative strand RNA viruses of the family Paramyxoviridae. The broad outlines of their biology and molecular genetics have been defined in extensive previous studies by this laboratory and others. The HPIV genome encodes a nucleoprotein N, phosphoprotein P, large polymerase protein L, internal matrix protein M, and fusion F and hemagglutinin-neuraminidase HN transmembrane surface glycoproteins. F and HN are the two viral neutralization antigens and the major protective antigens. In addition, the P gene encodes one or more accessory proteins (depending on the serotype) from one or more additional ORFs. Our present goal is to develop attenuated versions of HPIV1 and 3 that have been engineered to express the fusion F protein of human respiratory syncytial virus (RSV) from an added gene. This provides a bivalent vaccine candidate against the respective HPIV and RSV. Compared to RSV strains, the HPIVs replicate more efficiently in cell culture and do not share the physical instability of RSV strains. They also form spherical particles compared to the large filaments of RSV, making them more amenable to filtration and other steps in manufacture. These attributes would make RSV vaccines based on HPIV vectors much easier to manufacture, distribute, and use compared to attenuated RSV strains. These advantages may be essential for extending RSV vaccines to resource-challenged countries. Furthermore, in experimental animals, boosting responses to an earlier primary immunization with an attenuated RSV strain was more efficient using an HPIV/RSV vector as opposed to re-dosing with the attenuated RSV strain. Thus, HPIV/RSV vectors given subsequent (perhaps one year later) to an initial RSV immunization provide a unique way to achieve improved RSV boosts. We have been evaluating a number of parameters of vaccine vector design using, as proof of principle, an attenuated HPIV3 vaccine candidate called B/HPIV3. This virus consists of bovine PIV3 in which the F and HN genes have been replaced by those of HPIV3, yielding a chimeric virus that is attenuated in primates due to the bovine backbone, and which bears the neutralization and major protective F and HN antigens of HPIV3. Both the empty B/HPIV3 vector and B/HPIV3 expressing the unmodified RSV F protein were well-tolerated in infants and young children. However, the latter construct was poorly immunogenic for RSV F, and analysis of specimens of shed vaccine in nasal washes showed that 50% of specimens had evidence of loss of expression of RSV F. Therefore, we have been working to increase the immunogenicity and stability of the RSV F insert. We previously found that insertion of the RSV F gene into the first (pre-N) and second (N-P) gene positions of B/HPIV3 yielded up to a 69-fold increase in expression of RSV F compared to insertion at gene position 6. Expression of RSV F was further enhanced 5-fold by codon-optimization. In addition, RSV F was stabilized in the pre-fusion (pre-F) conformation (the conformation that is the most efficient in inducing neutralizing antibodies) by an added disulfide bond (called DS, a mutation developed by VRC/NIAID). In the hamster model, the DS form of RSV F induced an increased quantity and quality of RSV-neutralizing serum antibodies and increased protection against wt RSV challenge, compared to native F. High quality antibodies are those that neutralize RSV in vitro in the absence of added complement: remarkably, expression of unmodified F did not induce detectable high quality antibodies in hamsters, whereas the DS form induced a substantial titer. We further modified RSV F by replacing its cytoplasmic tail (CT) domain, or its CT plus transmembrane (TM) domains (TMCT), with its counterparts from BPIV3 F. This resulted in RSV F being packaged in the B/HPIV3 particle with an efficiency similar to that of RSV particles. Enhanced packaging was substantially attenuating in hamsters (10- to 100-fold) and rhesus monkeys (100- to 1000-fold). Nonetheless, TMCT-directed packaging substantially increased the titers of high quality serum RSV-neutralizing antibodies induced in hamsters and rhesus monkeys. In rhesus monkeys, the combination of packaging plus DS pre-F stabilization resulted in a 30-fold increase in serum RSV-neutralizing titers in the absence of added complement, respectively, compared to pre-F stabilization alone, despite the much-lower replication due to attenuation by TMCT. In the present report, the pre-F conformation was further stabilized by adding cavity-filling mutations to the DS form of F, resulting in the DS-Cav1 form (developed by VRC/NIAID). Also, codon-usage was modified to have a lower content of CpG dinucleotides, which can be inducers of innate immunity. This DS-Cav1 F ORF was evaluated in rB/HPIV3 in three forms: (i) DS-Cav1 ectodomain lacking the TMCT domains which would result in its secretion; (ii) DS-Cav1 without the TMCT packaging signal, and (iii) DS-Cav1 with the TMCT packaging signal. The DS-Cav1 ectodomain was efficiently secreted but was poorly immunogenic. Full-length DS-Cav1, with or without the vector-specific TMCT packaging signals, induced high titers of pre-F specific antibodies in hamsters and improved quality of RSV-neutralizing serum antibodies, compared to previous versions of RSV F. Codon-optimized RSV F containing fewer CpG dinucleotides had higher F expression, replicated slightly more efficiently in vivo, and was more immunogenic in hamsters as assessed by titers of high-quality RSV-neutralizing serum antibodies and protective efficacy against RSV challenge. Analysis of innate immune responses to the vectors in cell culture indicated that the reduced CpG content resulted in reduced type I and III interferon responses and a modest increase in vector replication. The combination of DS-Cav1 pre-F stabilization, optimized codon-usage, reduced CpG content, and vector packaging significantly improved vector immunogenicity and protective efficacy against RSV. This provides an improved HPIV3-vectored RSV vaccine candidate that will be manufactured into clinical trial maternal for pediatric clinical evaluation. The genetic stability of the RSV F insert in the B/HPIV3 vector was evaluated at several stages during optimization by a double-immunostaining plaque assay designed to detect expression of both RSV F and vector proteins. Surprisingly, the stability of RSV F expression typically was very high and usually was not affected by enhanced expression, or pre-F stabilization, or packaging via TMCT. An occasional preparation was found to have some loss of expression of RSV F, but this could simply be discarded and replaced by a replicate preparation. As noted, packaging the RSV F protein into the B/HPV3 vector increased its level of attenuation in rhesus monkeys 100- to 1000-fold, while providing increased quantity and quality of serum RSV-neutralizing antibodies. However, this reduction in replication is not needed for safety, since the B/HPIV3 vector already was satisfactorily attenuated on its own, and the further 100- to 1000-fold reduction in replication may unnecessarily reduce immunogenicity due to reduced antigen expression. Therefore, we also are inserting the TMCT version of RSV F into other backbones that are less restricted than B/HPIV3, and therefore should yield a construct that is not unnecessarily over-restricted for replication and therefore should be more immunogenic.